H.G.M. Leufkens

Medicines Evaluation Board, Netherlands, Utrecht, Utrecht, Netherlands

Are you H.G.M. Leufkens?

Claim your profile

Publications (364)867.12 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Breast cancer is the most common type of cancer among women worldwide. In low and middle-income countries (LMICs), appropriate selection of medicines on national essential medicines lists (NEMLs) is a first step towards adequate access to treatment. We studied selection of systemic treatments for breast cancer on NEMLs and assessed its alignment with treatment guidelines for different types of early and advanced breast cancer. Furthermore, influence of country characteristics on the selection was investigated. NEMLs from 75 LMICs were studied for inclusion of all components of therapy in each stage of breast cancer according to international consensus guidelines. The results were then grouped by income level, WHO region and the NEMLs' release date. Non parametric tests were used for statistical analysis. Unlike HER2-targeted therapies (<10 %), aromatase inhibitors (12 %) and taxanes (28 %); tamoxifen and first generation chemotherapeutic regimens (e.g., anthracycline-based regimens) were frequently found in the NEMLs (71-78 %). Consequently, all components of treatment for "Luminal A" early breast cancer and non HER2 overexpressed advanced breast cancer were found on the NEMLs of over 70 % of countries. However, 40 % of the low income countries did not have all the components of therapy for any type of early breast cancer in their NEMLs, and adequate treatment of HER2 overexpressed breast cancer was hardly possible with the current selections. Recent NEMLs were more aligned with the guidelines (p < 0.05). Eastern Mediterranean and African regions less frequently incorporated all components of breast cancer treatment in their NEMLs. Alignment of selection with guidelines' recommendations was inconsistent for different types of early and advanced breast cancer in NEMLs. Regular updates and more attention to clinical guidelines is therefore recommended.
    BMC Cancer 12/2015; 15(1):591. DOI:10.1186/s12885-015-1583-4 · 3.36 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Net Gain Analysis (NGA) is proposed as an alternative to Responders Analysis (RA) as a more comprehensive method to tap clinical relevance of the effect of treatment. NGA is the group difference in responders minus the group difference in deteriorators; while RA is the group difference in responders. We examined the performance of these two methods in a dataset consisting of individual patient data from 10 randomized controlled trials (N=2666) of five different antipsychotics in patients with acute mania by comparing the rank ordering of the five compounds according to both systems (NGA and RA). The rank order did not differ between the 2 systems but the inferiority of one compound was revealed more evidently by the NGA in comparison to the RA. Copyright © 2015. Published by Elsevier Inc.
    Regulatory Toxicology and Pharmacology 07/2015; 73(1). DOI:10.1016/j.yrtph.2015.06.020 · 2.03 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The probability of initiating with anti-osteoporosis therapy increased from 7 % in 2000 to 46 % in 2010. This improvement was greater for patients over the age of 75 years. Men, those overweight, having dementia or exposed to antipsychotics, sedatives/hypnotics or opioid analgesics were significantly less likely to receive anti-osteoporosis drugs. The objective of this study was to examine trends and determinants of anti-osteoporosis drug prescribing after hip fracture in the UK between 2000 and 2010. Data were extracted from the UK Clinical Practice Research Datalink for patients ≥50 years who had a first hip fracture between 2000 and 2010 and who did not currently (≤6 months prior) receive anti-osteoporosis drugs (bisphosphonates, strontium ranelate, parathyroid hormone, calcitonin and raloxifene) (n = 27,542). The cumulative incidence probability of being prescribed anti-osteoporosis drugs within 1 year after hip fracture was estimated by Kaplan-Meier life-table analyses. Determinants for treatment initiation were estimated by Cox proportional hazards models. The probability of being prescribed any anti-osteoporosis drug after hip fracture increased from 7 % in 2000 to 46 % in 2010. This trend was more marked in patients ≥75 years. The increase in prescribing of anti-osteoporosis drugs was complemented by a similar increase in vitamin D/calcium provision. Cumulative incidence of receiving anti-osteoporosis therapy was greater at any given point in time in women (8 % in 2000, 51 % in 2010) compared to men (4 % in 2000, 34 % in 2010). In addition to male gender, multivariable Cox regression identified reduced likelihood of receiving anti-osteoporosis drugs for those being overweight, having dementia and exposed to psychotropic drugs (antipsychotics, sedatives/hypnotics) or opioid analgesics. Although the prescribing of anti-osteoporosis drugs after hip fracture has increased substantially since 2000, the overall rate remained inadequate, particularly in men. With the continuing increase in the absolute number of hip fractures, further research should be made into the barriers to optimise osteoporosis management.
    Osteoporosis International 05/2015; 26(7). DOI:10.1007/s00198-015-3098-x · 4.17 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Long-term persistence with anti-osteoporosis drugs and determinants for discontinuation among fracture patients were examined. Persistence was 75.0 and 45.3 % after 1 and 5 years, respectively. Those aged ≥80 years were at increased risk of early discontinuation. Within 1 year after discontinuation, 24.3 % restarted therapy, yet 47.0 % persisted for 1 year. The risk of osteoporotic fracture can effectively be reduced with use of anti-osteoporosis drugs. However, little is known about persistence with these drugs after fracture where subsequent fracture risk is high. The aims were to determine long-term persistence with anti-osteoporosis drugs among fracture patients, including its determinants, and to describe restart and subsequent persistence. A cohort study was conducted within the Dutch PHARMO Database Network. Patients aged ≥50 years (n = 961) who received anti-osteoporosis drugs within 1 year after fracture, but not in the preceding year, were included (2002-2011). Persistence (defined as the proportion on treatment) and the proportion restarting after discontinuation were estimated using Kaplan-Meier analyses. Time-dependent Cox regression was used to identify determinants of non-persistence including age, sex, initial dosage regime, fracture type, comorbidities, and drug use. Persistence with anti-osteoporosis drugs was 75.0 % (95 % confidence interval (CI) 72.0-77.7) and 45.3 % (95 % CI 40.4-50.0) after 1 and 5 years, respectively. A significant determinant of non-persistence was age ≥80 years (reference 50-59 years: adjusted hazard ratio [adj. HR] 1.65; 95 % CI 1.15-2.38). This effect was not constant over time (≤360 days following initiation: adj. HR 2.07; 95 % CI 1.27-3.37; >360 days: adj. HR 1.08; 95 % CI 0.62-1.88). Within 1 year after discontinuation, 24.3 % (95 % CI 20.1-29.2) restarted therapy, yet 47.0 % persisted for 1 year. This study identified suboptimal persistence with anti-osteoporosis drugs among fracture patients. Major target groups for measures aimed to improve persistence may be those aged >80 years and those restarting therapy.
    Osteoporosis International 03/2015; 26(6). DOI:10.1007/s00198-015-3084-3 · 4.17 Impact Factor
  • S. Kleijnen · H.G.M. Leufkens · A. Boer · W. Goettsch · M. Fathallah
    Value in Health 11/2014; 17(7):A661-A662. DOI:10.1016/j.jval.2014.08.2428 · 3.28 Impact Factor
  • British Journal of Clinical Pharmacology 10/2014; 78(4):764-765. · 3.88 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Risk Management Plans (RMPs) have become a cornerstone in pharmacovigilance of new drugs in Europe. The RMP was introduced in 2005 to support a proactive approach in gaining knowledge on safety concerns, through early planning of pharmacovigilance activities. The rate at which uncertainties in the safety profile are resolved through this proactive approach is, however, unknown. We therefore examined the evolution of safety concerns in the RMP after initial approval for a selected cohort of 48 drugs, to provide insight into the knowledge gain over time. We found that 20.7% of the uncertainties existing at approval had been resolved five years after approval. Since new uncertainties were included in the RMP at a similar rate, their overall number remained approximately equal. The relatively modest accrual of knowledge, as demonstrated in this study through resolution of uncertainties, suggests that opportunities for optimization exist while ensuring feasible and risk proportionate pharmacovigilance planning.Clinical Pharmacology & Therapeutics (2014); Accepted article preview online 15 September 2014. doi:10.1038/clpt.2014.184.
    Clinical Pharmacology &#38 Therapeutics 09/2014; 96(6). DOI:10.1038/clpt.2014.184 · 7.90 Impact Factor
  • European geriatric medicine 09/2014; 5:S62. DOI:10.1016/S1878-7649(14)70126-3 · 0.73 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The risk of a subsequent major or any fracture after a hip fracture and secular trends herein were examined. Within 1 year, 2.7 and 8.4 % of patients sustained a major or any (non-hip) fracture, which increased to 14.7 and 32.5 % after 5 years. Subsequent fracture rates increased during the study period both for major and any (non-hip) fracture. Hip fractures are associated with subsequent fractures, particularly in the year following initial fracture. Age-adjusted hip fracture rates have stabilised in many developed countries, but secular trends in subsequent fracture remain poorly documented. We thus evaluated secular trends (2000-2010) and determinants for the risk of a subsequent major (humerus, vertebral, or forearm) and any (non-hip) fracture after hip fracture. Patients a parts per thousand yen50 years with a hip fracture between 2000 and 2010 were extracted from the UK Clinical Practice Research Datalink (n = 30,516). Incidence rates, cumulative incidence probabilities, and adjusted hazard ratios (aHRs) were calculated. Within 1 year following hip fracture, 2.7 and 8.4 % of patients sustained a major or any (non-hip) fracture, which increased to 14.7 and 32.5 % after 5 years, respectively. The most important risk factors for a subsequent major fracture within 1 year were the female gender [aHR 1.90, 95 % confidence interval (CI) 1.51-2.40] and a history of secondary osteoporosis (aHR 1.54, 95 % CI 1.17-2.02). The annual risk increased during the study period for both subsequent major (2009-2010 vs. 2000-2002: aHR 1.44, 95 % CI 1.12-1.83) and any (non-hip) facture (2009-2010 vs. 2000-2002: aHR 1.80, 95 % CI 1.58-2.06). The risk of sustaining a major or any (non-hip) fracture after hip fracture is small in the first year. However, given the recent rise in secondary fracture rates and the substantial risk of subsequent fracture in the longer term, fracture prevention is clearly indicated for patients who have sustained a hip fracture.
    Osteoporosis International 07/2014; 25(11). DOI:10.1007/s00198-014-2799-x · 4.17 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Drug utilization studies have applied different methods to various data types to describe medication use, which hampers comparisons across populations. The aim of this study was to describe the time trends in antidepressant prescribing in the last decade and the variation in the prevalence, calculated in a uniform manner, in seven European electronic healthcare databases. Annual prevalence per 10,000 person-years (PYs) was calculated for 2001-2009 in databases from Spain, Germany, Denmark, the United Kingdom (UK), and the Netherlands. Prevalence data were stratified according to age, sex, antidepressant type (selective serotonin re-uptake inhibitors [SSRIs] or tricyclic antidepressants [TCAs]) and major indications. The age- and sex-standardized prevalence was lowest in the two Dutch (391 and 429 users per 10,000 PYs) and highest in the two UK (913 and 936 users per 10,000 PYs) populations in 2008. The prevalence in the Danish, German, and Spanish populations was 637, 618, and 644 users per 10,000 PY respectively. Antidepressants were prescribed most often in 20- to 60-year-olds in the two UK populations compared with the others. SSRIs were prescribed more often than TCAs in all except the German population. In the majority of countries we observed an increasing trend of antidepressant prescribing over time. Two different methods identifying recorded indications yielded different ranges of proportions of patients recorded with the specific indication (15-57 % and 39-69 % for depression respectively). Despite applying uniform methods, variations in the prevalence of antidepressant prescribing were obvious in the different populations. Database characteristics and clinical factors may both explain these variations.
    European Journal of Clinical Pharmacology 05/2014; 70(7). DOI:10.1007/s00228-014-1676-z · 2.97 Impact Factor
  • Source
    S Pouwels · A de Boer · H G M Leufkens · W E J Weber · C Cooper · T P van Staa · F de Vries
    [Show abstract] [Hide abstract]
    ABSTRACT: The aim of this study was to evaluate fracture risk in patients with Guillain-Barré syndrome (GBS). No association with risk of fracture was observed for GBS patients compared with controls. Only GBS patients using pain treatment had a doubled risk of fracture. Symptoms of Guillain-Barré syndrome (GBS) may vary from mild difficulty in walking to complete paralysis, which may increase the risk of fractures. Therefore, the aim of this study was to evaluate fracture risk in patients with GBS. We conducted a retrospective cohort study using the UK Clinical Practice Research Datalink (1987-2012). Each patient with GBS was matched by year of birth, sex, and practice, up to six patients without a history of GBS. Outcome measure was any fracture. There were no associations between GBS and any fracture, adjusted hazard ratio (AHR) 1.01 (95 % confidence interval [CI] 0.77-1.33), or osteoporotic fracture, AHR 0.76 (95 % CI 0.50-1.17), compared with controls. Stratification to gender, age, and duration since diagnosis did not show an association either. Only for GBS patients using pain treatment, risk of fracture was doubled AHR 1.97 (95 % confidence CI 1.21-3.21) compared with controls. The risk of fracture in GBS patients exposed to pain treatment was equivalent to risk of fracture among controls exposed to pain treatment. No association with risk of fracture was observed for GBS patients compared with controls. Only GBS patients using pain treatment had a doubled risk of fracture, but their risk was equivalent to fracture risk among controls exposed to pain treatment.
    Osteoporosis International 04/2014; 25(7). DOI:10.1007/s00198-014-2705-6 · 4.17 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: High cancer mortality rates in low- and middle-income countries (LMICs) have raised concerns regarding access to oncology medicines. Essential medicines are those which satisfy the primary health care needs and provide a basis for public procurement or reimbursement decisions in LMICs. We explored selection of oncology medicines in LMICs through investigating national essential medicines lists (NEMLs) for cancer treatments. Recently updated NEMLs were retrieved for 76 countries. Oncology medicines were classified based on therapeutic categories. Countries were clustered based on geographic regions, income levels and burden of cancer (mortality and morbidity). Indicators of frequency (number of medicines), diversity (number of therapeutic (sub)categories) and more importantly absence were measured and compared across countries using parametric and nonparametric tests. The overall median number of oncology medicines on NEMLs was 16 (interquartile range = 23) chosen predominantly from subcategories of 'antineoplastic agents', with substantial variation across regions and income groups. Five countries did not select any oncology medicine and 68% did not have any 'hormones and related agents' on their NEMLs. Newer technologies like targeted therapies were infrequently incorporated. The cluster of countries suffering most from the burden of cancer selected more essential oncology medicines and diversified further. The observed selection of oncology essential medicines can reflect insufficiencies and inequalities in access to cancer treatments at least in the public sector of LMICs. Further resources need to be allocated from governments and international organizations to tackle the problem of access to oncology medicines in these countries.
    Annals of Oncology 01/2014; 25(1):270-6. DOI:10.1093/annonc/mdt514 · 7.04 Impact Factor
  • European geriatric medicine 09/2013; 4:S178. DOI:10.1016/j.eurger.2013.07.596 · 0.73 Impact Factor
  • Source
    E. Beers · A.C.G. Egberts · H.G.M. Leufkens · P.A.F. Jansen
    European geriatric medicine 09/2013; 4(8):S178. DOI:10.1016/j.eurger.2013.07.595 · 0.73 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The aim of the study was to determine fracture risk in incident muscular dystrophy (MD) patients. Patients with MD are at a 1.4-fold increased risk of fracture as compared with population-based control patients. Risk further increased among elderly and female patients and among patients exposed to oral glucocorticoids. Muscular dystrophies (MDs) are inherited diseases causing muscle weakness and thereby increase the risk of falling and detrimental effects on bone. Both are recognised risk factors for fracture. Therefore, the aim of this study was to determine the hazard ratio of fracture in patients with MD. We conducted a retrospective cohort study using the UK General Practice Research Database (1987-2012). Each patient with MD was matched by year of birth, sex and practice to up to six patients without a history of MD. Outcome measure was all fractures. As compared with control patients, risk of any fracture was statistically significantly increased in MD patients (adjusted hazard ratio [AHR], 1.40; 95 % confidence interval [CI], 1.14-1.71). An increased risk of fracture was observed among MD patients with female gender (AHR, 1.78; 95 % CI, 1.33-2.40) and an increasing age as compared with control patients. Stratification to Duchenne MD showed no association with fracture, whereas risk of fracture was increased twofold among patients with myotonic dystrophy (AHR, 2.34; 95 % CI, 1.56-3.51). MD patients had an almost tripled risk of fracture when they used oral glucocorticoids in the previous 6 months as compared to non-users with MD. Patients with MD are at a 1.4-fold increased risk of fracture as compared with population-based control patients. Especially in older age groups and female gender, the fracture risk of MD versus non-MD patients is increased, whereas exposure to glucocorticoids further increased fracture risk among MD patients.
    Osteoporosis International 08/2013; 25(2). DOI:10.1007/s00198-013-2442-2 · 4.17 Impact Factor
  • Source
    J.L. Boer · E. Beers · T.C. Egberts · H.G.M. Leufkens · P.A.F. Jansen
    Clinical Therapeutics 08/2013; 35(8):e89-e90. DOI:10.1016/j.clinthera.2013.07.266 · 2.73 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Marketing authorization application dossiers of 17 orphan drugs (ODs) and 51 non-ODs evaluated by the European Medicines Agency (EMA) in the period 2009-2010 were compared. We aimed to identify whether any differences existed between ODs and non-ODs in number and type of deficits brought forward during the EMA review, regarding the clinical development plan, clinical outcome and medical need and studied whether these deficits were similarly associated with marketing approval in the EU. In 71% of the ODs dossiers and 65% of the non-ODs dossiers marketing approval was granted. Differences in deficits were found, but similarities in the way ODs and non-ODs were reviewed and marketing approval decisions were taken, underline that regulatory standards are equally high.
    Drug discovery today 07/2013; 18(19). DOI:10.1016/j.drudis.2013.06.012 · 6.69 Impact Factor
  • Source
    H A van den Ham · O H Klungel · H G M Leufkens · T P van Staa
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Anticoagulation control is often summarised using the percentage of time spent in therapeutic range (TTR). This method does not describe the timing and severity of fluctuations in the International Normalised Ratio (INR). Objective: To evaluate whether the TTR method can be improved by considering the patterns of INR over time. Methods: The cohort included adults aged 40+ years with atrial fibrillation (AF) and laboratory records of INR as recorded in the UK Clinical Practice Research Datalink. Statistical clustering techniques based on simple INR measures were used to describe the patterns of INR. Nested case-control studies calculated the odds ratios (ORs) for the risk of stroke, bleeding and mortality with TTR and different INR patterns. It was also evaluated whether cluster analyses improved the prediction of outcomes over TTR. Results: 27,381 patients were studied with a mean age of 73. The OR for mortality was 3.76 (95% confidence interval [CI] 3.03-4.68) in patients with <30% TTR compared to patients with 100% TTR. INR patterns were found to be best described by six different clusters. The cluster with the most unstable pattern was associated with the largest risk of mortality (OR 10.7, 95% CI 8.27-13.85) and stroke (OR 3.42, 95% CI 2.08-5.63). INR measures that predicted death independent of TTR included absolute difference between two subsequent INR measurements and change relative to the mean over time. Conclusion: Cluster analysis of INR patterns improved prediction of clinical outcomes over TTR and may help to identify warfarin users who need additional anticoagulation monitoring. © 2012 International Society on Thrombosis and Haemostasis.
    Journal of Thrombosis and Haemostasis 10/2012; 11(1). DOI:10.1111/jth.12041 · 5.72 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Objective Assessment of the effectiveness compared with alternative treatment(s) plays an important role in many jurisdictions in determining the reimbursement status of pharmaceuticals. This type of assessment is often referred to as a relative effectiveness assessment (REA) and is carried out by many jurisdictions. Increased sharing of information across jurisdictions may save costs and reduce duplication. The objective of this study was to explore the main similarities and differences in the major methodological aspects of REA in multiple jurisdictions.Methods Data were gathered with a standardized data extraction form by searching publicly available information and by eliciting information from representatives at relevant organizations.ResultsOf the initially included 35 jurisdictions, data were gathered for 29 jurisdictions. There seem to be substantial similarities on the choice of the comparator, the role of indirect comparisons, and preferred end points in REAs (except for the use of health state utilities). Jurisdictions, however, differ in whether effectiveness (usual circumstances of health care practice) is estimated in case no (comparative) effectiveness data are available and how this is done.Conclusion Some important methodological aspects for REA are approached in a similar way in many jurisdictions, indicating that collaboration on assessments may be feasible. Enhanced collaboration in the development of methods and best practices for REA between jurisdictions will be a necessary first step. Important topics for developing best practice are indirect comparisons and how to handle the gap between efficacy and effectiveness data in case good quality comparative effectiveness data are not yet available at the time of reimbursement decisions.
    Value in Health 09/2012; 15(6):954-960. DOI:10.1016/j.jval.2012.04.010 · 3.28 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Achtergrond: Veel mensen met angst- of slaapstoornissen krijgen van hun dokter een benzodiazepine voorgeschreven. Langdurig gebruik zorgt echter voor gewenningsproblemen en hoge maatschappelijke kosten. Om deze effecten te beperken heeft het ministerie van VWS de vergoeding van benzodiazepinen per 1 januari 2009 uit de basisverzekering gehaald. Sindsdien betalen de meeste gebruikers de kosten van benzodiazepinen zelf. Methode: In de gegevens van het Landelijk Informatie Netwerk Huisartsenzorg (LINH) keken wij hoeveel mensen tussen januari 2007 en december 2009 voor het eerst de diagnose slaap- en angststoornissen kregen bij de huisarts. Daarna onderzochten wij hoeveel recepten voor een benzodiazepine deze patiënten kregen en hoeveel van hen er ook weer mee stopten, en zo ja wanneer. Ook keken wij naar het aantal recepten voor een selectieve serotonineheropnameremmer (SSRI). Resultaten: Van de 13.596 patiënten die aan de voorwaarden voldeden had iets meer dan de helft een angststoornis, de anderen hadden een slaapprobleem. In 2009 kregen minder mensen de diagnose slaap- of angststoornis te horen dan in 2008. Ook het benzodiazepinegebruik daalde. Van de patiënten met slaapstoornissen kreeg 67,0% in 2008 een benzodiazepine voorgeschreven, versus 59,1% in 2009, en voor angststoornissen waren deze percentages 33,7% in 2008 versus 30,1% in 2009. Ook het aantal mensen dat meer dan één recept kreeg, daalde: voor slaapproblemen van 42,6% naar 35,0%, voor angststoornissen van 42,6% naar 36,4%. Slechts een klein (maar toenemend) aantal angstpatiënten stapte over op een SSRI: 0,3% in 2009 tegenover 0,1% in 2008. Conclusie: Het reduceren van de vergoeding heeft geleid tot een lichte daling in het aantal diagnoses, voorschriften en vervolgrecepten. Deze daling was groter bij slaapproblemen dan bij angststoornissen. De uitkomsten ondersteunen het beleid van de regering en laten zien dat er ruimte bestaat om het gebruik van deze geneesmiddelen te reduceren. (aut.ref.)
    Huisarts en wetenschap 07/2012; 55(7). DOI:10.1007/s12445-012-0144-1

Publication Stats

7k Citations
867.12 Total Impact Points


  • 2013–2015
    • Medicines Evaluation Board, Netherlands
      Utrecht, Utrecht, Netherlands
  • 1993–2014
    • Utrecht University
      • Division of Pharmacoepidemiology and Pharmacotherapy
      Utrecht, Utrecht, Netherlands
  • 1997–2012
    • Netherlands Institute for Space Research, Utrecht
      Utrecht, Utrecht, Netherlands
  • 2011
    • Diakonessenhuis Utrecht
      Utrecht, Utrecht, Netherlands
  • 2008
    • Medicines and Healthcare products Regulatory Agency (MHRA)
      Londinium, England, United Kingdom
  • 2003
    • Pharmo Institute for Drug Outcomes Research
      Utrecht, Utrecht, Netherlands
    • Altrecht GGZ
      Utrecht, Utrecht, Netherlands
    • Atrium Medisch Centrum Parkstad
      Heerlen, Limburg, Netherlands
  • 2002–2003
    • St. Elisabeth Ziekenhuis Tilburg
      • Department of Surgery
      Tilburg, North Brabant, Netherlands
    • University Medical Center Utrecht
      • Julius Center for Health Sciences and Primary Care
      Utrecht, Utrecht, Netherlands
  • 1999–2002
    • Maastricht University
      • Department of Epidemiology
      Maestricht, Limburg, Netherlands
  • 1995
    • Erasmus Universiteit Rotterdam
      • Department of Internal Medicine
      Rotterdam, South Holland, Netherlands